On Nov 21, 2012, at 11:07 AM, Nick Kledzik <kledzik at apple.com<mailto:kledzik at apple.com>> wrote: On Nov 21, 2012, at 8:55 AM, Relph, Richard wrote: AMD would like to add new functionality to ranlib (and later ar and nm) and to the bits of LLVM Core that read (and later write) archives. Herewith a terse summary of the change, which we want to improve support of OpenCL for multiple GPUs in a single run-time. Conceptually, a serialized archive is really 2 pieces: a few header members and a set of normal file members. There are no constraints on the normal members in the 'pure' archive format. They could be text files, pictures, or, as we're all familiar with, object modules. Most object file archives are "libraries" and the have a special header member that is a global symbol table, associating global scope names with defining object module members in the archive body. We have N very large archives, defining essentially the same set of symbols. Many of the normal file members of each are duplicated in other archives, but not all. The goal is the produce a single "super-archive" that contains 1 copy of each unique object file member no matter how many archives it is part of, and N symbol table members representing each of the original N archives. Let me see if I understand your need here. You are dynamically generating code and need to link it. The linking step requires some support routines which makes sense to have in a static library. Since this must work on machines not set up with developer tools, you are packaging the static library inside a DLL/DSO. In addition, with all the minor variations of GPUs, having a separate archive for every GPU type would be too large, so you need some way to remove duplicates of support functions. Close. But it isn't "support routines". It's the entire OpenCL "built-in" library… thousands and thousands of functions, some tiny, some large, some merely aliases. If the above summary is close, then here are two other ideas that avoid the need for archive/TOC changes: 1) Have lots of little archives which removes duplicates. Give each archive a unique name, then have a lookup table which lists which sequence of archives to use for which specific GPU. When linking for a particular GPU, you pass the linker that particular sequence of little archives to search. This is close to what we have now, but find unworkable because the Linker doesn't take a "set" of libraries. So if library A calls something in B calls something in A, you end up having to make multiple passes over the libraries, which isn't efficient and creates other headaches. This is what we are trying to 'solve' by going to a single library, but we want to preserve the space advantage of the current approach. 2) Have one giant archive for per GPU family and use name mangling scheme to filter. For instance, the compiler emits references to support routine "foo" as "foo$gpu1". Then you construct the support libraries with aliases for each support function. So a particular implementation of foo in an archive may show up in the symbol table through its aliases "foo$gpu3", "foo$gpu4", "foo$gpu7". When the linker is only looking for "foo$gpu1", it will ignore all other foo implementations and just pick the one aliased to "foo$gpu1". I'll have to think about this… We try to treat the IR emitted by the front-end as pretty family/device neutral to the extent that's possible. We'd want to implement this as a pass between the linking of user-produced code and the subsequent linking of the OpenCL library. And we'd obviously have to create a tool that takes the current library members and aliases them to the 'correct' unique implementation. BTW, Apple/Darwin has a similar issue with supporting multiple CPUs. Our solution is "fat" archives. The ranlib tool sorts all archive members by cpu, builds "thin" archive libraries for each cpu, then concatenates the thin libraries together with a "fat header" which specifies the file offset and size of each thin archive. The linker, when it comes across a fat static library, looks at the fat header thens seeks along to the contained thin archive relevant to the cpu type being linked. This scheme works well, but it can produce large files because there is no attempt to reduce duplicates. Sounds pretty similar in terms of the problem… except we are motivated to avoid duplicates because a lot of the members are truly identical, whereas the "fat" archive probably has few identical members at the binary level because of instruction set and calling convention differences. This is most similar to my first considered approach. Thanks, Richard -Nick The symbol table for each original archive can properly index to the relevant members in the archive, even if other members in the super-archive (not referenced in this particular symbol table, of course) define the same symbols. I've considered 3 approaches to the problem so far. All involve a new archive member type. First, a new archive member type "up front" that describes each of the original archives and its symbol table. Second, a normal/default symbol table member "up front" and a new archive member type that describes alternate symbol tables contained in the archive. Third, a "hiding" archive member type that is essentially a way to "skip over" additional normal archive file headers to reach the first normal member, which (in all approaches) all archives share. The third, I think, requires the least changes to the existing implementation, so I'm leaning towards it. The "hiding" archive member would have the "file name" of the represented archive immediately following the member header, followed by a completely normal archive representation starting with "<!arch>\n" and optionally including an additional "hiding" archive member covering even more hidden archives. The plan is to extend the Archive class to provide for a way for clients to select a desired archive. I also will enhance ranlib to accept multiple archive names on the command line and produce the "super-archive" from ranlib. A further need we have is to serialize the TOCs and the super-archive in a memory image (our archives are embedded in our DLL/SO, not stored separately on disk) and then provide an interface to the relevant LLVM classes (Linker, primarily) for accessing archives in memory rather than on disk, a feature absent from the current implementation. For our purposes, extending the Archive class to support specification of the archive using a memory object instead of a file, recognizing the "hiding" member type, and extending ranlib to produce the new super archives is all we really need. Any thoughts or suggestions would be welcome. Thanks, Richard _______________________________________________ LLVM Developers mailing list LLVMdev at cs.uiuc.edu<mailto:LLVMdev at cs.uiuc.edu> http://llvm.cs.uiuc.edu<http://llvm.cs.uiuc.edu/> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20121122/0e802f18/attachment.html>
On Nov 21, 2012, at 4:00 PM, Relph, Richard wrote:> On Nov 21, 2012, at 11:07 AM, Nick Kledzik <kledzik at apple.com> wrote: >> On Nov 21, 2012, at 8:55 AM, Relph, Richard wrote: >>> AMD would like to add new functionality to ranlib (and later ar and nm) and to the bits of LLVM Core that read (and later write) archives. >>> Herewith a terse summary of the change, which we want to improve support of OpenCL for multiple GPUs in a single run-time. >>> >>> Conceptually, a serialized archive is really 2 pieces: a few header members and a set of normal file members. There are no constraints on the normal members in the 'pure' archive format. They could be text files, pictures, or, as we're all familiar with, object modules. Most object file archives are "libraries" and the have a special header member that is a global symbol table, associating global scope names with defining object module members in the archive body. >>> >>> We have N very large archives, defining essentially the same set of symbols. Many of the normal file members of each are duplicated in other archives, but not all. The goal is the produce a single "super-archive" that contains 1 copy of each unique object file member no matter how many archives it is part of, and N symbol table members representing each of the original N archives. >> Let me see if I understand your need here. You are dynamically generating code and need to link it. The linking step requires some support routines which makes sense to have in a static library. Since this must work on machines not set up with developer tools, you are packaging the static library inside a DLL/DSO. In addition, with all the minor variations of GPUs, having a separate archive for every GPU type would be too large, so you need some way to remove duplicates of support functions. > > Close. But it isn't "support routines". It's the entire OpenCL "built-in" library… thousands and thousands of functions, some tiny, some large, some merely aliases. > >> If the above summary is close, then here are two other ideas that avoid the need for archive/TOC changes: >> 1) Have lots of little archives which removes duplicates. Give each archive a unique name, then have a lookup table which lists which sequence of archives to use for which specific GPU. When linking for a particular GPU, you pass the linker that particular sequence of little archives to search. > > This is close to what we have now, but find unworkable because the Linker doesn't take a "set" of libraries. So if library A calls something in B calls something in A, you end up having to make multiple passes over the libraries, which isn't efficient and creates other headaches. This is what we are trying to 'solve' by going to a single library, but we want to preserve the space advantage of the current approach.The gnu linker has the option: --start-group archives --end-group which tells the linker to keep searching the list of archives (not just make one pass). The darwin linker does this by default. BTW, I had assumed you were using the "system linker". But that won't work if you are adding some new kind of archive table of contents. I guess that means you have some in-process code that provides linking functionality? If so, I'd like to understand this better to see if this is something lld <http://lld.llvm.org/> might want to support some day. -Nick -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20121124/b4b122f0/attachment.html>
On Nov 24, 2012, at 2:48 PM, Nick Kledzik <kledzik at apple.com<mailto:kledzik at apple.com>> wrote: On Nov 21, 2012, at 4:00 PM, Relph, Richard wrote: On Nov 21, 2012, at 11:07 AM, Nick Kledzik <kledzik at apple.com<mailto:kledzik at apple.com>> wrote: On Nov 21, 2012, at 8:55 AM, Relph, Richard wrote: AMD would like to add new functionality to ranlib (and later ar and nm) and to the bits of LLVM Core that read (and later write) archives. Herewith a terse summary of the change, which we want to improve support of OpenCL for multiple GPUs in a single run-time. Conceptually, a serialized archive is really 2 pieces: a few header members and a set of normal file members. There are no constraints on the normal members in the 'pure' archive format. They could be text files, pictures, or, as we're all familiar with, object modules. Most object file archives are "libraries" and the have a special header member that is a global symbol table, associating global scope names with defining object module members in the archive body. We have N very large archives, defining essentially the same set of symbols. Many of the normal file members of each are duplicated in other archives, but not all. The goal is the produce a single "super-archive" that contains 1 copy of each unique object file member no matter how many archives it is part of, and N symbol table members representing each of the original N archives. Let me see if I understand your need here. You are dynamically generating code and need to link it. The linking step requires some support routines which makes sense to have in a static library. Since this must work on machines not set up with developer tools, you are packaging the static library inside a DLL/DSO. In addition, with all the minor variations of GPUs, having a separate archive for every GPU type would be too large, so you need some way to remove duplicates of support functions. Close. But it isn't "support routines". It's the entire OpenCL "built-in" library… thousands and thousands of functions, some tiny, some large, some merely aliases. If the above summary is close, then here are two other ideas that avoid the need for archive/TOC changes: 1) Have lots of little archives which removes duplicates. Give each archive a unique name, then have a lookup table which lists which sequence of archives to use for which specific GPU. When linking for a particular GPU, you pass the linker that particular sequence of little archives to search. This is close to what we have now, but find unworkable because the Linker doesn't take a "set" of libraries. So if library A calls something in B calls something in A, you end up having to make multiple passes over the libraries, which isn't efficient and creates other headaches. This is what we are trying to 'solve' by going to a single library, but we want to preserve the space advantage of the current approach. The gnu linker has the option: --start-group archives --end-group which tells the linker to keep searching the list of archives (not just make one pass). The darwin linker does this by default. BTW, I had assumed you were using the "system linker". But that won't work if you are adding some new kind of archive table of contents. I guess that means you have some in-process code that provides linking functionality? If so, I'd like to understand this better to see if this is something lld <http://lld.llvm.org/> might want to support some day. We are using the Linker class, which uses the Bitcode/Archive class, which is why I was focusing on that class to support the new functionality. Once I have Bitcode/Archive enhanced, and llvm-ar generates the new "multi-archives", I was going to add a method to the Linker class to support linking "in-memory archives"… not strictly part of the Archive enhancements, but something we else we need. Richard -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20121206/26fce778/attachment.html>